Should Your Agency Offer a Results Guarantee? How to Choose, Price and Word One

Sooner or later a prospect on a sales call asks the question every agency owner hears: "And what if it doesn't work?" You can answer with case studies, with a slide about your process, or with a promise. The promise is the answer that closes, which is why so many owners end up asking whether their agency should guarantee results. Then they search for advice and find a wall of articles explaining that guaranteed results are a scam.

Both sides are half right. Most of the articles warning against guarantees are written for buyers who are judging vendors that promise rankings on keywords nobody searches. The question an agency owner is asking is a different one. It is not "can I promise the result is certain?" because nobody can. It is "should I carry the cost if the result doesn't happen, and can I afford to?" That question has a real answer, and it depends on what your work controls, how quickly your results appear, and what a miss would cost you.

This guide is written for the agency side of the table. It covers the two tests that decide whether a results guarantee makes sense for your service, the four structures you can actually honor, how to price the risk before you print it, how to word it so a non-expert owner understands it in one read, and where it belongs in your outreach and on your sales calls.

The strongest offer in the market, described in the wrong words

One SEO agency owner we onboarded had already built a guarantee most of his competitors would not dare offer. If a client's site did not reach page one, the setup fee came back, and the monthly retainer was not billed at all until the client ranked. He was carrying almost all of the risk himself.

His outreach described all of that in four words: "pay per result SEO." That is the name the industry uses for the pricing model. To the owner of a plumbing company or a dental practice it means very little. It does not say what the result is, when it arrives, or what happens to their money if it doesn't. The guarantee existed. Nobody reading his messages could see it.

Rewritten in plain words, the same offer became "we get you on page one or you don't pay." Nothing about the economics changed, but a busy owner understands it in one read. That is the whole job of a guarantee: to make the buyer's worst case small and obvious. A guarantee that needs explaining is only doing half of that job.

So agency guarantees go wrong in two ways: some agencies offer one they cannot afford, and others already carry the risk and get no credit for it. The rest of this guide deals with both.

A guarantee moves risk; it does not create results

Start with what a guarantee actually does. It does not make good results more likely. Your campaign performs the same whether or not you promised anything. What a guarantee changes is who pays when the results miss. Without one, the client absorbs the loss. With one, you do.

That shift is worth making when you can carry the risk more cheaply than the buyer can, and usually you can, for one reason: you know your miss rate and they do not. You have delivered this service before. You know roughly how often it works, how long it takes, and what the failures had in common. The buyer is guessing, and a buyer who is guessing assumes the worst. A guarantee replaces their guess with your track record, with the price of the occasional miss built in.

Two tests before you offer any guarantee

Test one: does your work control the number?

Guarantee only what your own work determines. An agency that books appointments can guarantee appointments, because booking is its work. It cannot honestly guarantee closed revenue, because the client runs the sales call. A web designer can guarantee a launch date and a page speed score. It cannot guarantee that the new site doubles enquiries, because enquiries depend on traffic, on the offer, and on how quickly the client answers the phone.

SEO sits in the middle, which is why it attracts so much criticism. You control the work. Google controls the ranking, and Google's own guidance on hiring an SEO says plainly that nobody can guarantee a number-one position. What an SEO agency can do is guarantee page-one results on searches it has already won for similar clients, in a defined area, inside a window long enough to be realistic. That is a promise based on your track record, not on Google's cooperation.

A useful rule: if hitting the number depends on something the client does after you hand over your work, the guarantee belongs one step earlier in the chain.

Test two: does the result arrive before the buyer's patience runs out?

A guarantee also needs a window, and the window has to fit how your results actually appear. Paid ads usually show within the first few weeks whether a campaign produces leads. Outreach shows replies within days and booked calls within the first month. SEO often takes months to move. A guarantee window shorter than the time your results need is a promise you will break on schedule.

When your results are slow, you have two honest options. Guarantee an earlier milestone you can measure sooner, such as technical fixes finished and pages published by a set date. Or lengthen the window and delay the client's spend to match it, which is what the SEO agency above did by not billing the retainer until the rankings arrived.

Four guarantee structures you can actually honor

Once a guarantee passes both tests, the next decision is what happens when you miss. There are four common answers, and they cost you very different amounts.

1. Refund the setup fee or first payment. The cleanest version, and the easiest for a buyer to understand. If you miss, you lose the refund plus the work you already did. It suits services where the first month is mostly your own labor and your hit rate is high.

2. Defer billing until the result arrives. You do not charge the monthly fee until the number is hit. A miss costs you cash flow more than a refund, because you are funding the work for however long it takes. It suits slow results like SEO, where you are confident the result will come but cannot say exactly when.

3. Keep working free until you hit it. Often called a make-good. A miss costs your labor plus any hard costs you cover. Buyers value it less than they value a refund, because it does not return money already spent, and it only helps a client who intends to stay.

4. Let them leave without penalty. No minimum term, or unused months refunded on cancellation. A miss costs you future revenue rather than past revenue. It is the weakest promise, but it is an honest one for services where no outcome can be guaranteed at all.

You can stack structures. The SEO agency above stacked the first two: a refunded setup fee and a deferred retainer. Stacking shrinks the buyer's worst case almost to nothing, but it also multiplies your exposure, so it only makes sense with a high hit rate. Paying only for delivered results is the same idea taken to its end point, and it is how pay-per-appointment lead generation works: the provider carries the whole risk and prices it into each appointment.

If none of the four fit because your outcome depends too heavily on the client, guarantee what you do instead of what happens: response times, delivery dates, reporting on a fixed day. These promises are cheap to keep, and they still answer the buyer's real worry, which is often "will they disappear once I've paid?"

Price the risk before you print it

A guarantee is an insurance policy you are selling, and every insurer starts with two numbers: how often the bad thing happens, and what it costs when it does.

Call the share of clients who would trigger your guarantee your miss rate. Call what one miss costs you the miss cost: the refund, the unpaid months, or the cash tied up while billing is deferred. Every client you sign under the guarantee carries an expected cost of miss rate times miss cost.

Now the other side of the ledger. A guarantee earns its keep by closing more of the sales calls you already have. The test is simple to state:

The extra clients the guarantee wins, times what a client is worth to you over their lifetime, has to be larger than the number of guaranteed clients you sign, times your miss rate, times the miss cost.

A worked example with round ratios rather than prices. Say you hold forty sales calls a quarter and close one in five, so eight clients. A guarantee lifts that to one in four, so ten clients. Your miss rate is one in ten, so you expect to pay out on one of the ten. The guarantee won you two clients and cost you one payout. It pays whenever two clients are worth more than one miss cost, which for a setup refund against months of retainer is almost always true. Now change one number: make the miss rate one in three. You expect three or four payouts against the same two extra clients, and the answer depends heavily on which structure you chose. A refunded setup fee probably still pays. Several months of deferred billing per miss may not.

Two lessons follow. The miss rate matters as much as the size of the refund, and it is the number you control most, through which clients you accept. And if you do not know your miss rate, you cannot run the sum at all. The close-rate lift is the other unknown; rather than guess it, offer the guarantee on your next batch of proposals and compare against the batch before. What a normal close rate looks like, and what moves it, is covered in average close rates on sales calls.

Write the conditions before the headline

Every honest guarantee has conditions, and careful buyers read reasonable conditions as a sign that the guarantee is real. Write them before you write the headline, because the conditions decide what you are actually promising. Four things need defining.

• What counts. Which searches, in which area, checked with which tool. What counts as a lead or an appointment: does it have to be in the client's target market, and does it count if the prospect doesn't show?
• When the clock starts. From when the work goes live, not from the day the contract is signed. If onboarding depends on the client sending you access or content, say so.
• What the client must do. Approve content within a set number of days, provide access, attend booked calls, and keep the target market stable during the window. These are not escape hatches. They are the inputs the result depends on.
• What the remedy is, and who decides. The exact refund or credit, when it is paid, and how a disagreement about whether the guarantee triggered gets settled.

These are the same four questions a careful buyer asks about any vendor's guarantee, which we lay out from the buyer's side in appointment-setting guarantees explained. Answer them in writing before the prospect has to ask, and the guarantee stops being a negotiation.

The line between a fair condition and an escape hatch is specificity. "Client approves content within five business days" is a condition. "Subject to client cooperation" is an escape hatch, and buyers can tell the difference. This is general information rather than legal advice; have whoever reviews your contracts check the final wording before it goes into one.

Say it in words the buyer would use

Back to "pay per result SEO." The phrase failed because it describes a pricing model from the seller's side. Buyers do not think in pricing models. They think: what do I get, by when, and what happens to my money if I don't get it?

A guarantee that a non-expert owner understands in one read has three parts, in that order: an outcome they can see, a timeframe, and what happens to their money. A few patterns:

• SEO: "Page one on Google for the three searches we agree on, within the window we set together, or your setup fee comes back, and you pay nothing monthly until you're there."
• PPC: "If your campaign hasn't brought in the number of leads we agree on by the end of month one, the next month's management is free."
• Web design: "Your new site is live by the date in your contract, or that phase of the project is on us."
• Bookkeeping: "Your books are closed by the same day every month, or that month's fee is waived."

Notice what none of these say. There is no "guaranteed results," no "risk-free," no percentage. The vaguer the promise, the more it reads as marketing and the less it reassures anyone. The specific version is also the honest one. "Page one for searches we agree on in advance" answers the main criticism of SEO guarantees, which is that agencies pick easy keywords nobody searches. When the searches are named and agreed together, that criticism no longer applies.

Where the guarantee belongs: outreach, the call, the contract

In outreach, near the top

If your guarantee is your strongest asset, it belongs in the first message, not on page four of a proposal. Most cold messages describe a service. Very few describe a worst case the buyer no longer has to worry about. A one-line version of your guarantee is often a better reason to reply than any claim about your skill. How to pair it with an ask that actually gets answered is covered in what to ask for in a cold outreach message.

On the call, as the answer to "what if it doesn't work?"

Don't open the call with it. Open with the prospect's problem and how you would fix it. A guarantee lands hardest as the answer to the objection, after the prospect has seen the plan and is weighing the risk. Explain it in two sentences, then explain the conditions as the things that make the result happen: "We need your sign-off on content within five days, because that is what keeps us on schedule for page one." Put that way, conditions sound like a plan rather than small print.

In the contract, word for word

The written guarantee should match what was said on the call, as close to word for word as you can make it. Disputes start when the call promised "page one" and the contract says "improved visibility." If a salesperson other than you runs the calls, give them the exact sentence to use, and keep it to the sentence.

When a guarantee is the wrong move

There are situations where a guarantee costs more than it earns.

• The service is new. No track record means no miss rate, so you would be betting your margin on a number you have never measured. Run it with a handful of clients on short terms first, and add a guarantee once you know how often it works.
• The result depends on the client. If the number you would guarantee is decided by their sales team, their pricing or their follow-up, move the guarantee earlier in the chain or don't offer one.
• The client insists on it under pressure. A prospect who needs a guarantee before anything else is sometimes a prospect for whom one more failed vendor would be a crisis. The clients who most need the guarantee are often the ones most likely to trigger it.
• Your cash flow cannot carry it. Deferred billing is generous and expensive. If three slow clients at once would strain payroll, choose a structure with a capped, known cost.
• It is covering for a weak offer. If you need the guarantee to close every deal, the problem is upstream: the niche, the price, or the offer itself. It is worth checking whether your market can comfortably afford you, which we cover in which niches can afford your agency.

A guarantee is also not permanent. Agencies often use one to win their first clients and retire it once case studies can do the same job. Taking it away is one of the steps for raising your agency prices, and like any other price step it is best tested on new clients before existing ones. If you are earlier than that and have no results to show yet, a guarantee is one of several ways to shrink a buyer's worst case, alongside shorter commitments and a smaller first scope; we compare them in how to get clients without case studies.

How we apply the same tests

TaskBlink's own guarantee is at least 3 booked appointments in your first 30 days, or you don't pay. It is built around the two tests in this guide. It guarantees appointments, which our work controls, rather than closed deals, which depend on your sales call. And thirty days fits how quickly outreach produces booked calls. We find businesses that match your ideal customer, confirm every phone number is a real working cell, and run the outreach by text, email and phone, so more of your calendar holds the conversations your own guarantee then has to convert. For marketing agencies, that means more interested buyers hearing a well-worded guarantee.

Frequently asked questions